Research On Position Detection And Compression Technology Of The EAST Plasma Based On High-speed CCD

In tokamak thermonuclear reaction process, the processes of plasma breakdown and current formation are very complex and the induction eddy current on the device wall will interfere in the measured signal seriously. The measurement data obtained in this situation can not react the actual situation of plasma. Fast CCD image sensor is now playing a pivotal role in plasma diagnostic technique research due to its fast response speed and good imaging results. With a good visibility, CCD unfolds the radiation phenomena of plasma to the researchers in real time and recognizes the position of plasma current accurately and quickly. Therefore, CCD can be used as an important tool for the detection of plasma configuration and will also provide effective data support to start and feedback control plasma.According to the structural characteristics of EAST device and the test requirements, a fast CCD video image acquisition system was built in this paper with 4M180 camera of IO Industries company as the core component. The system can achieve the related settings of camera and the acquisition card, for example the resolution, the sampling rate, the exposure time and so on. In addition, the system is not only able to complete the tasks of collecting, transmitting and storing images, but also there are two different acquisition modes including trigger and continuous acquisition. In this paper, the hardware and software design architecture of the imaging system are also described in detail.The plasma position recognition technology has been studied in this paper based on the above image acquisition system. The position of the plasma does have great significance in the steady running of tokamak as well as the smooth completion of the nuclear fusion reaction. According to the characteristics of the plasma discharge image, the improved Canny algorithm was used in this paper. The local standard deviation was selected to be the reference value when determining the scale parameter σ of Canny algorithm. At the same time, threshold automatic acquisition method based on OTSU was selected to determine the high and low threshold value of Canny operator. After the boundary position of the plasma was obtained, curve fitting was performed to the obtained image by least square method so that the center position of the plasma could be obtained. Finally, the actual position of plasma in the tokamak device was obtained by comparing the plasma position coordinates with the calibration image.In addition, in order to make the acquired image data be quickly transferred and stored in the large capacity data server in a real-time, the plasma image was compressed based on compressed sensing theory. The original image can be reproduced with high SNR. According to the principles of compressed sensing, data sparsification was achieved, observation matrix was built and image compression was completed. Finally, the image was reproduced with the OMP algorithm and the effect is significant.